https://github.com/ctu-vras/monoforce - latest

- Jupyter Notebook
Published by RuslanAgishev 12 months ago

https://github.com/ctu-vras/monoforce - T-RO 2025 submission

The repository version for the paper submitted to the IEEE Transactions on Robotics (T-RO) journal. The corresponding arXiv paper: https://arxiv.org/abs/2502.10156

- Jupyter Notebook
Published by RuslanAgishev about 1 year ago

https://github.com/ctu-vras/monoforce - v0.1.1

• The pre-trained models predict not only the terrain shape, but also friction values.
• New Terrain Encoder models are added: LidarBEV and BEVFusion. The architectures allow to use lidar scans as an additional sensor input for terrain properties prediction.

- Jupyter Notebook
Published by RuslanAgishev about 1 year ago

https://github.com/ctu-vras/monoforce - Checkpoints

Terrain Encoder weights trained on ROUGH and RELLIS-3D datasets. The PyTorch checkpoint names have the following format: {model}_{dataset}_{robot}.pt

- Jupyter Notebook
Published by RuslanAgishev over 1 year ago

https://github.com/ctu-vras/monoforce - ICML 2024 DiffXYZ

ICML 2024 DiffXYZ

The package version accepted to the Differentiable Almost Everything Workshop: https://differentiable.xyz/

- Jupyter Notebook
Published by RuslanAgishev over 1 year ago

https://github.com/ctu-vras/monoforce - IROS-2024 submission

- Jupyter Notebook
Published by RuslanAgishev almost 2 years ago

https://github.com/ctu-vras/monoforce - Geometric (lidar) height map prediction

LSS model capable to replicate height map from Lidar sensor from input images.

- Jupyter Notebook
Published by RuslanAgishev about 2 years ago